NetApp is backed into a corner

Reuters indicates that EMC will up its bid for Data Domain to as much as $35 per share. As previously posted, Data Domain’s products will fit easily into EMC’s product line replacing EMC’s current Quantum-based appliances. With this increased offer, EMC is increasing the pressure on NetApp and reaffirming their commitment to acquire Data Domain.

What does this mean?

EMC is stronger financially then NetApp as this table illustrates:

$ Millions EMC NetApp
Numbers as of 12/31/08 4/24/09
Revenue $14,876 $2,153
Operating Income $1,576 $47
Cash & Equivalents $5,844 $1,494

Given the situation above, if EMC really wants Data Domain they will prevail. EMC has the financial strength to make an offer that NetApp cannot match. The only question is “Is EMC really committed to acquiring Data Domain?” In my opinion (and that of others I have spoken to), EMC wants Data Domain, and the Reuters pieces confirms this.

EMC’s aggressiveness puts NetApp in a tough spot. Their smaller cash position means that they cannot offer a competitive cash-only option unless they find new sources of cash such as a loan. Securing a line of credit is difficult in these tight credit markets and would likely leave NetApp at a disadvantage if they acquire Data Domain.

I have heard rumors that NetApp might try to play the regulatory card and pursue an anti-trust case against EMC if they acquire Data Domain. This will be a difficult to justify especially since NetApp claims market leadership in deduplication. According to this press release, NetApp has 37,000 systems running deduplication while the latest analyst reports suggest that Data Domain has shipped between 3,000 and 3,500 systems.

In summary, EMC has backed NetApp into a corner. NetApp has two options at this point and both are less than ideal:

  1. All out bidding war for Data Domain – There is no guarantee that NetApp can win, and if they did, at what cost? They would be strapped with debt and left in a difficult financial position. Obviously this is not favorable to long-term competitiveness and would put them at a disadvantage to EMC.
  2. Cede Data Domain to EMC – From a financial standpoint, this is probably the better option, but it is not what Data Domain and NetApp’s management wants. It also would leave NetApp without a compelling deduplication solution and in a weakened competitive position against a combined EMC/Data Domain.

I have watched this bidding war unfold with interest. With the latest salvo from EMC, it is clear that they will do what it takes to acquire Data Domain. NetApp cannot effectively compete with EMC’s stronger financial position and should start thinking about plan B.

5 replies on “NetApp is backed into a corner”

If the bidding is going to continue, then this has clealy become more ego driven than business or technolgy driven.

I am actually more interested in seeing how the “loser” of this bidding war will respsond. It will be very difficult for the loser, whoever it may be, to go back to the status quo and claim convincingly that they have the best inline target based dedeulication solution.

I can see them, particulary if it is EMC that loses out, possibly turning to post-processing target based dedeupe as an alternative. Of course, it is also possible that HP, Hitachi, and other competitors may see post processing as the alternative solution to combat a NetApp/Data Domain or EMC/Data Domain competitor.

In any case, that could make things very interesting for FalconStor or your company, Sepaton. I see many suitors in the horizon.

Hi and thank you for your comment,

I do not doubt that this war could become more ego driven than business driven; however, I do believe that NetApp has a much stronger business need to acquire Data Domain than EMC. (More in an upcoming post.) Unfortunately, NetApp has much shallower pockets. The “loser” of this bidding war will be put in a tough spot, and I think that NetApp has more to lose than EMC.

I disagree with your characterization of post-process target deduplication. You appear to be positioning inline and post-process deduplication as separate solutions, and suggest that vendors must now look at post-process solutions. This is a common misconception. Customers or vendors need to think about the ability to meet backup windows, recovery time and ROI objectives. It is not simply a question of post-process or inline, but rather how a given solution can meet these requirements. Any entity looking to acquire a deduplication solution is best served to look at these metrics rather than focus on the post-process/inline debate.

I agree that we are living in interesting times!

Thanks for your response.

My point is not that post-processing is better than inline, but that whoever does not get Data Domain may have to align themselves with the post-processing camp as a defense against the winning suitor. After all, if NetApp “wins”, EMC can’t really backtrack and say that their Quantum inline solution is superior; but they could align themselves with FalconStor, for example, and “claim” superiority with post-processing instead.

As you know, many of these technical decisions have little to do with pure technology considerations.

I see your point and agree that the business decisions often take precedence over technology considerations. However, I still slightly disagree with your point. The “loser” will choose a deduplication solution and claim superiority over Data Domain regardless of whether it is an inline or post-process system. I do not believe that they will be forced to choose a post-process approach to differentiate from Data Domain’s inline technology.

This raises another point which is that the “loser” will have a hard time claiming technical superiority for their existing solution after haven just aggressively bid for Data Domain.

In an amusing twist on all this, EMC already has Falconstor technology (the VTL part) in their EDL solution. They could have just gone with Falconstor de-dupe as well, and made their overall solution much cleaner and simpler, but they apparently looked over Falconstor’s de-dupe offering and said “No thanks”, opting instead for Quantum on the de-dupe side.

I’m reminded of the words of the Old Knight in “Indiana Jones and the Last Crusade” about picking the chalice… “Choose wisely”… and what happened to the villain who “chose poorly”.

It just feels like EMC “chose poorly” with Quantum, and they’re paying dearly for doing so, although it’s unlikely that they’ll collapse into dust anytime soon. If they merely drive up the price for DDUP and then let Netapp have it, I agree that EMC will still have to do something different, having tacitly admitted that they have crap for de-dupe right now.

The loser in this bidding war will simply have to do something else, make some kind of major change or acquisition, since they’ve both basically admitted that they have an inferior de-dupe play, regardless how much they hyped it over the past 12-24 months. It doesn’t much matter which way they go, in-line, post-process, or combo, as long as they change it up.

So who’s out there to provide a quick injection of de-dupe secret sauce? Falconstor (which EMC should have just done before, frankly, but having passed once on it, cannot really do now with any credibility), Sepaton, Copan, Symantec’s Puredisk… possibly others.

Copan uses Falconstor through and through (like EMC should have), for VTL and de-dupe functions. So there’s no licensing their secret sauce, it would come down to buying them and replacing existing overlapping products. Possible but ugly for either of them.

Symantec’s Puredisk is its own technology from the ground up, and Symantec has many OEM agreements to show that it is fine with licensing its technology to other players, generally with some kind of non-compete agreement as part of the deal. Puredisk is also very different from other VTL-style de-dupe, but competes well against them. Possible advantageous match here.

Sepaton… to me, this is the obvious play for the loser, although I have no idea what the buy-out price might be, or if it would be worth it. Frankly, Sepaton has a great story, once you get past the “post process” mental block, allowing it to do “wire speed” restore as well as ingest. It scales to huge sizes, and provides better HA and higher speeds than competing solutions. With post-process, you need more disk up front, but that’s the only way you can get the speeds they deliver, both for backups and for restores.

I’d watch for the loser to cast an acquisitive eye on Sepaton, or possibly get in bed with Symantec.

Film at 11…

Leave a Reply

Your email address will not be published. Required fields are marked *